Data Interview Qs - Selecting computers for a company

Suppose you're trying to help a company (with hundreds of employees) determine which computers to purchase. The company has been able to pull utilization data by employee that classifies users into 3 bins, depending on how much they use their computer in their work:

• Low usage - spends a lot of time in meetings, checking email, doing people management
• Average usage - requires some compute power, with balanced mix of heads down/technical work along with a good amount of meetings/email writing
• High usage - power user, relies heavily on computer performance

Additionally, they've surveyed employees to collect the relative importance of the following variables describing a computer's performance:

• Memory
• Processing
• Storage
• Price inverse - this metric was given to you by the company, with the directive that price should have a 25% weight in the purchase decision.

The results of the survey data can be found here.

Lastly, the company is looking to purchase a maximum of 3 different computer models, and have compiled the following list scoring their memory, processing, storage, and relative price. Each dimension is scored from 0-10, with 10 being the best.

Given this information, provide the company with a recommendation on which computers to purchase. Be sure to lay out the logic behind your recommendation and the assumptions/steps you took to arrive there. You can consider this more of an open-ended/case study question.

To help get you started, below is code (Python) to import the referenced datasets:

    
# Import libraries
    
%matplotlib inline
    
import pandas as pd
    
import matplotlib.pyplot as plt
    
import numpy as np
    
# Import data
    
employee_utilization = pd.read_csv('https://raw.githubusercontent.com/erood/interviewqs.com_code_snippets/master/Datasets/computer_case/employee_utilization.csv')
    
computer_options = pd.read_csv('https://raw.githubusercontent.com/erood/interviewqs.com_code_snippets/master/Datasets/computer_case/computer_options.csv')
    
survey_results = pd.read_csv('https://raw.githubusercontent.com/erood/interviewqs.com_code_snippets/master/Datasets/computer_case/survey_results.csv')
    

Our solution will be written using Python for premium users, and lay out our assumptions/logic behind our recommendation as we would to a stakeholder at the company.